Running HAFS suite in regional mode with Ferrier-Aligo microphysics

[ericaligo-NOAA]

The HAFS suite with the Ferrier-Aligo (FA) microphysics runs in global mode when running the RT, however, it fails in a regional configuration. In debug mode, the error points to line 119 in Dycore_typdefs.F90 (see below). Line 119 is the allocation of surface pressure (see below). Any idea what could be going on here? Is this a namelist issue? I'm running off of GFS ICs/LBCs. The FA microphysics doesn't have a separate cloud ice array (qi) and instead has a total ice (qs) and also has a diagnostic 3D rime factor array that is initialized to a value of 1 in the model code. The GFS ICs/LBCs do not have total ice or the rime factor and it's not clear to me if this could be causing problems near the boundaries.

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
ufs_model 0000000009F89CFC Unknown Unknown Unknown
libpthread-2.17.s 00002B2011CC1630 Unknown Unknown Unknown
libiomp5.so 00002B201041440F Unknown Unknown Unknown
libiomp5.so 00002B2010414EFC Unknown Unknown Unknown
ufs_model 0000000009FC49A6 Unknown Unknown Unknown
ufs_model 000000000585F969 dycore_typedefs_m 119 DYCORE_typedefs.F90

!--- surface pressure
allocate (Coupling%p_srf (IM))
Coupling%p_srf = clear_val

[climbfuji]

Based on my previous investigation of this issue it seems that the stack is corrupted. Is the allocate (Coupling%p_srf (IM)) statement the first allocate statement ? If yes, you could try to swap this with a following allocate statement (including the initialization of course) - if it then fails in the other (=first) one, then you can be pretty certain that it is not specific to p_srf but hints to a corrupted stack.